Scientific Data Analysis using Collaborative Virtual Reality
ABSTRACT. Over decades of experience at the UC Davis W.M. Keck Center for ActiveVisualization in the Earth Sciences (KeckCAVES), collaborative virtualreality has proven itself as a powerful medium for interactive analysisof scientific data.
In this presentation, we will discuss the example of how KeckCAVES VRsoftware was used during landing site selection for the 2012 Curiositymission, and how the experience gained during that and other projects isinforming our current efforts to develop a VR framework supporting awide range of collaborative applications running on low-cost commodityVR hardware.
Anomaly Detection in Spacecraft Telemetry: Forecasting vs. Classification
ABSTRACT. Anomaly detection in spacecraft telemetry is critical for the success and safety of space missions. Traditional methods often rely on forecasting and threshold techniques to identify anomalies [1]–[5]. This paper presents a comprehensive comparison of traditional forecast-based anomaly detection against two innovative classification methods, including a direct classification and an image classification through Gramian Angular Field (GAF) transforms [6], which have only been analysed in other domains but not for spacecraft anomaly detection. All our investigated systems leverage deep learning architectures and use the popular real SMAP/MSL spacecraft data from [2]. Our findings suggest that direct classification provides a marginal but statistically significant improvement in anomaly detection over traditional methods. However, image classification, while less successful, offers promising directions for future research. The study aims to guide the selection of appropriate anomaly detection techniques for spacecraft telemetry and contribute to the advancement of automated monitoring systems in space missions.
Standards-Based Development and Design for High-Performance Single Board Computers
ABSTRACT. The presentation will discuss standards-based board development and design for the high-performance single board computers. An overview will be provided highlighting the advantages of standards-based SBC designs. These capabilities are essential in providing system integration simplification in increasingly complex systems.
Total Ionizing Dose Radiation Testing of NVIDIA Jetson Orin NX System on Module
ABSTRACT. In space applications, the adoption of commercial-off-the-shelf (COTS) single-board computers (SBCs) is increasingly favored due to their size, weight, and power (SWaP) efficiency. This study addresses the critical need for understanding the radiation tolerance of such devices within low earth orbit (LEO) space missions, where intense ionizing radiation presents a substantial risk to electronic component functionality. Focusing on the NVIDIA Jetson Orin NX, a leading COTS SBC, we evaluated the radiation resilience of both its 8GB and 16GB models under total ionizing dose (TID) conditions. Our investigation reveals significant consistency in radiation tolerance among the models tested, surviving past 36.20 krad(Si). This underscores the considerable resilience to the effects of radiation and the absence of performance degradation despite challenges related to thermal management. These findings are crucial for the aerospace community, informing the deployment of COTS SBCs in environments with high radiation exposure and impacting considerations for mission success and device longevity.
DEMO: Scientific Data Analysis using Collaborative Virtual Reality
ABSTRACT. This is a VR demo of the system presented in the associated talk. Abstract follows:
Over decades of experience at the UC Davis W.M. Keck Center for Active
Visualization in the Earth Sciences (KeckCAVES), collaborative virtual
reality has proven itself as a powerful medium for interactive analysis
of scientific data.
In this presentation, we will discuss the example of how KeckCAVES VR
software was used during landing site selection for the 2012 Curiosity
mission, and how the experience gained during that and other projects is
informing our current efforts to develop a VR framework supporting a
wide range of collaborative applications running on low-cost commodity
VR hardware.
Tutorial Part 1: Introduction to Robotic Simulations in NVIDIA Isaac Sim & Omniverse
ABSTRACT. One day, everything that moves will be autonomous. Robotic automation has made significant strides forward, driven by advancements in hardware and artificial intelligence capabilities that have opened new avenues in simulation and strive for autonomy. This workshop we will give a technical introduction to the Omniverse and Isaac SIM platforms, a cutting-edge solution for robotics and simulation.
We will start off with a generic presentation section to introduce use-cases, value and vision of the platform and some examples on how it can be applied to the space industry. Next we'll move over to a more technical hands-on lab where you'll dive into the simulation loop of a 3D engine, learning to initialize experiments with objects, robots, and physics logic, and build some small robotics control tasks and applications within the simulation environment .
The hands-on piece is a technical beginner level, and thus you don't need any prior knowledge on Isaac SIM, apart from basic python understanding.
Some additional packages might be required if using Linux for the Omniverse Streaming Client (details in the documentation)
Note: We will use the NVIDIA Deep Learning Institute platform for the hands-on portion of this workshop. Attendees will be handed a personal code during the workshop, that will give them access to one of the self-paced paid courses. The codes will be shared during the workshop. These are personal, and can only be redeemed to one specific course only. You can find more information on how to redeem the DLI platform codes in the attached pdf.
Additionally, course content and access to the environment will be given for up to 1 year after the workshop. There are also other self-paced courses available for further learning.
ABSTRACT. The emerging paradigm of neuromorphic computing has the potential to provide high-performance, low-power computing for edge artificial intelligence. This promise of high performance and low power consumption makes neuromorphic computing devices attractive for platforms at the edge: those that are constrained in size, weight, and power. Spacecraft fall into the category of edge platforms. In space, computing devices are subject to radiation effects not present on Earth, including single event effects (SEE) and total ionizing dose (TID). In this paper, we present the results of performing proton SEE testing and TID testing on an exemplar neuromorphic processor, the Intel Loihi.
Upgrading the Basics - Review of Micropac Offering
ABSTRACT. Extending from components to modules to plug in cards, power and data transport systems have new options. New product offerings are emerging from a renewed look at the needs of today’s missions and today’s mission assurance approaches.
Characterization of Single Event Functional Interrupts in COTS and Rad-Hard ARM Microcontrollers
ABSTRACT. This work evaluated the single event functional interrupt (SEFI) response of the commercial-off-the-shelf (COTS) and radiation hardened Cortex-M4 (M4) microcontrollers with the Armv7-M ISA. The microcontrollers were exposed to 200 MeV protons. The COTS M4 was further evaluated under carbon ions and alpha particles to assess if the control bits that depend on the software could have a significant influence on the SEFI cross section for this ISA. The SEU results show that both microcontrollers can experience multiple-cell upsets (MCUs), which could facilitate the accumulation of multiple-bit upsets (MBUs). MBUs could be a concern even for radiation hardened systems with ECC, because ECC crashes the system (SEFI) to correct MBUs.
SCS3740 & HPSC Rad-Hard SBC, DDR4, eMMC and Terabit NAND for Space
ABSTRACT. Flight proven and planned rad hard and rad tolerant single board computers, supporting memory technologies including DDR4 and high density managed NAND components.
AMD XQR Versal Adaptive SoCs Enable Next-Generation Signal Processing in Space
ABSTRACT. In this presentation we review qualification and radiation data for the AMD XQR Versal adaptive SoC devices, and look at how their reconfigurable heterogeneous computing resources enable demanding applications such as digital beamforming and STAP radar processing to be performed on orbit.
Tutorial Part 2: Introduction to Robotic Simulations in NVIDIA Isaac Sim & Omniverse
ABSTRACT. One day, everything that moves will be autonomous. Robotic automation has made significant strides forward, driven by advancements in hardware and artificial intelligence capabilities that have opened new avenues in simulation and strive for autonomy. This workshop we will give a technical introduction to the Omniverse and Isaac SIM platforms, a cutting-edge solution for robotics and simulation.
We will start off with a generic presentation section to introduce use-cases, value and vision of the platform and some examples on how it can be applied to the space industry. Next we'll move over to a more technical hands-on lab where you'll dive into the simulation loop of a 3D engine, learning to initialize experiments with objects, robots, and physics logic, and build some small robotics control tasks and applications within the simulation environment .
The hands-on piece is a technical beginner level, and thus you don't need any prior knowledge on Isaac SIM, apart from basic python understanding.
Some additional packages might be required if using Linux for the Omniverse Streaming Client (details in the documentation)
Note: We will use the NVIDIA Deep Learning Institute platform for the hands-on portion of this workshop. Attendees will be handed a personal code during the workshop, that will give them access to one of the self-paced paid courses. The codes will be shared during the workshop. These are personal, and can only be redeemed to one specific course only. You can find more information on how to redeem the DLI platform codes in the attached pdf.
Additionally, course content and access to the environment will be given for up to 1 year after the workshop. There are also other self-paced courses available for further learning.
ABSTRACT. Building towards a sustained lunar and deep-space presence requires advances in space infrastructure. Technologies commonly used on Earth, such as computer displays, cannot be naively incorporated into flight systems due to reliability concerns. Radiation-tolerant crew displays represent a critical technology gap NASA aims to address as part of its Moon to Mars roadmap. To date, crew displays lag significantly behind state-of-the-art terrestrial systems. Constrained by environmental challenges, legacy hardware, and safety requirements, current crew displays lack the ability to deliver high-resolution graphics thus restricting visual communication capabilities and crew autonomy. Moreover, the varying requirements of a diverse collection of surface and orbital lunar assets further complicate finding a generalized solution and design. These limitations and challenges necessitate the development of enhanced avionic systems to support future crewed missions. In this paper, we present key design considerations, a methodology, and a preliminary architecture to realize a radiation-tolerant display system. Balancing radiation tolerance, compute capability, and scalability, this paper describes a methodology to optimize the performance and reliability of a display computing system. Leveraging a hybrid design of commercial-off-the-self (COTS) and radiation-hardened (rad-hard) processors and components, a preliminary architecture is presented that includes hardware and software mitigation strategies that consider both cumulative and acute radiation effects. Lastly, a prototype is presented for benchmarking and validation.